Description of problem: I have a local printer which I wish to share on my LAN via IPP. I configure it with system-config-printer. Including sharing. Other machines do not see the queue. Note: this worked in FC3. When I look in /etc/cups/cupsd.conf, the subnet CIDR block is mentioned nowhere (unlike in FC3). Based on the FC3 version, the following things seem to be missing: [for each queue] Allow from 192.139.70.64/255.255.255.192 [Global, at end] BrowseAddress 192.139.70.127 Listen 192.139.70.122:631 The output of "printconf-tui --Xexport" on the FC4 system does mention the LAN subnet. I have attached a copy. Version-Release number of selected component (if applicable): system-config-printer-0.6.131.3-1 How reproducible: Every time that I've tried. Steps to Reproduce: 1.system-config-printer 2.select queue 3.Action:Sharing... 4.Add... 5.Network address 6.fill in network address and mask 7.OK 8.OK 9.Apply Actual results: other machines cannot see the queue /etc/cups/cupsd.conf does not reflect the network specified Expected results: otehr machines can see the queue /etc/cups/cupsd.conf reflects the network specified Additional info: Since system-config-printer overwrites /etc/cups/cupsd.conf every time, it seems pointless to hand-edit the config file.
Created attachment 125054 [details] output of printconf-tue --Xexport
Created attachment 125055 [details] cupsd.conf generated
As a work-around, I tried adding sharing for anything on Network devices eth0. This too made NO change to cupsd.conf. Then I gave up and said "All hosts". That seems to have caused changes to cupsd.conf. Remote printing seems to work. Queue changes percolated very slowly between the machines (something to do with "browsing", I think).
Sounds like rhpl.ethtool isn't working right. Please change the settings back to 'Network devices eth0', and try this as root: echo '#!/usr/bin/python' > /tmp/printconf-backend cat /usr/sbin/printconf-backend >> /tmp/printconf-backend chmod a+x /tmp/printconf-backend /tmp/printconf-backend --force-rebuild What does it say?
==== Output: ==== Rebuild forced on command line Add 'clawlex' configuration to /etc/cups/printers.conf Add 'clawlex' configuration to /etc/cups/cupsd.conf Create /etc/cups/ppd/clawlex.ppd PPD generated. Adjusting options. Set page margins: *ImageableArea Letter/US Letter: "36 36 576 756" For interfaces: ['@IF(eth0)'] ipaddr: ['192.139.70.88'] broadcast: ['192.139.70.127'] ==== end of output ==== Interesting: looks reasonable. The /etc/cups/cupsd.conf reflects this too. Stupidly, I didn't take a snapshot after I went back to eth0 but before issuing the commands you requested, so I don't know what fixed the /etc/cupsd.conf. The first line of /usr/sbin/printconf-backend is "#!/usr/bin/python -O" so your command sequence effectively just suppressed this -O. Do you suspect that python optimization is a problem? For what it is worth, I'm using Fedora Core's python-2.4.1-2 for x86_64.
The '-O' just suppresses these debugging messages. In theory, you just did what pressing 'Apply' would have done, except for restarting the CUPS daemon. Strange that it didn't work before.
Given the last couple messages, this sounds irreproducible. I suppose I'll have to close it, then. Also note that FC4 is no longer supported except for Fedora Legacy, and s-c-p has been completely rewritten for FC6.